search

ashawkey / stable-dreamfusion

Text-to-3D & Image-to-3D & Mesh Exportation with NeRF + Diffusion.
Apache License 2.0
7.99k stars 710 forks source link

Zero123 on multiple images #341

Open Bhavay-2001 opened 10 months ago

Bhavay-2001 commented 10 months ago

Hello authors.

I can run zero123 on a single image. However if I want to run it on a dataset or on multiple images, I need to create a .csv file for it. I have 2 questions regarding the same -

  1. How can I create the csv file if I don't know about the ratio, polar and azimuth of the image.
  2. If I want to run zero123 for multiple images which are not of same category. For example - car, shoes, bus, sofa etc. How can I create a csv file for the same and will the model be able to produce results for all of these?
honeyandme commented 5 months ago

Have you solved it? I also have the same question, can these three parameters be calculated by software?

Bhavay-2001 commented 5 months ago

Hi @honeyandme, I have stopped working on the project. However, if you are working on a research project, I would be happy to help you out with it. Thanks

joshkiller commented 3 months ago

Hi @Bhavay-2001 I'm currently working on my master thesis project since 2 weeks ago on 3D generative AI. and I would like any help from you.